| title | emoji | colorFrom | colorTo | sdk | sdk_version | app_file | pinned |
|---|---|---|---|---|---|---|---|
LoL Patch Update Chat-bot |
🏆 |
green |
indigo |
streamlit |
1.39.0 |
chatbot.py |
false |
An interactive chatbot that provides users with real-time access to the latest patch updates for League of Legends.
🚀 Live Demo: League Patch ChatBot
- RAG Chatbot: Google API for LLMs & text embedding, LangChain
- Vector Database (Patch Info Storage): Pinecone
- Automated Data Flow (Crawling & Storing Patch Data): Prefect, BeautifulSoup
- Database (Logging Full Chatbot Interactions for Future Optimization): Weights & Biases (WandB)
- User Interface: Streamlit
Simply visit the Live Demo and start chatting!
# Clone the repo
git clone https://github.com/ntdat02092002/LoL-ChatBot
cd LoL-ChatBot
# Install dependencies
pip install -r requirements.txtYou will need API keys from the following services:
Create a .env file or export the following environment variables:
WANDB_API_KEY=<your_wandb_api_key>
PINECONE_API_KEY=<your_pinecone_api_key>
GOOGLE_API_KEY=<your_google_api_key>
PINECONE_INDEX_NAME=<your_pinecone_index_name>
PREFECT_API_KEY=<your_prefect_api_key>
PREFECT_API_URL=<your_prefect_api_url>For setting up and deploying on your local machine, please refer to the official guide here using pinecone_data_flow.py script.
python prefect_deployment.py
prefect deployment run 'pinecone-flow/lolchatbot-data-flow'streamlit run chatbot.pyVisit the provided localhost link in your browser to access the chatbot UI.
For those who want to experiment with different LLMs or text embedding models (e.g., OpenAI, Claude, etc.), check out LangChain Documentation for model integrations.
To modify the chatbot:
- Edit
config.jsonfor model settings - Update
main.pyandpinecone_data_flow.pyaccordingly
✅ Automatically crawls detailed patch notes from the official game site to ensure up-to-date information.
✅ RAG-based retrieval model using LangChain with Pinecone as the vector database.
✅ Implements self-query retriever for more accurate and context-aware responses.
✅ Stores complete user interaction history under the hood for analysis and continuous improvement.
🔹 Optimize prompt templates for improved response quality
🔹 Implement memory & follow-up conversation capabilities
🔹 Expand chatbot scope beyond latest patch updates (include patch history & general game information)
💡 Contributions & Feedback: Feel free to open issues or PRs to improve this project!